It seems like, instead of trying to figure out all of the unknowns, it might be better to just make the assumption that problematic exotic plants will not be unknown. So if we go with this assumption, the unknown plants will likely be a combination of mostly native plants, and not currently problematic exotics. Thus, most of the plots here are exotic vs notexotic.
Main takeaways so far:
First, species counts. So, we see again the same pattern (I am just including 1m, plot, and site-level scales). The main thing here that jumps out at me is the messiness we were talking about seems like it might turn out to be a pretty interesting result, and first plot, illustrating the point that at continental scales the broader exotic paradigm bears out, but then zooming in closer everything’s super messy.
We can also look at the effect of relative cover of exotics - same messy site-to-site effect, but then the general relationship is clear
So now were getting into the alpha diversity, and from here I’ll just show the broad relationships.
So I guess here, we might be able to use aridity, elevation and latitude to model the NERRs, and maybe after removing the effects of covariates we’d get cleaner lines. That’s the next thing I was thinking about getting to. I might use one of those global gridded soil datasets to add in a few general soil variables like percent sand.
Variable importance plots
So I also have neondiveRsity creating beta diversity indexes now.
So now, we have neondiveRsity counting the number of families, and then also calculating shannon diversity using families instead of species.